A unified Smith predictor based on the spectral decomposition of the plant
|
|
- Sandra Chapman
- 5 years ago
- Views:
Transcription
1 A unified Smith predictor based on the spectral decomposition of the plant Qing-Chang Zhong and George Weiss Dept. of Electrical & Electronic Engineering Imperial College London Exhibition Rd., London SW7 2BT, UK Abstract We point out a numerical problem in the well-known modified Smith predictor and propose a unified Smith predictor (USP) which overcomes this problem. The proposed USP combines the classical Smith predictor with the modified one, after a spectral decomposition of the plant. We then derive an equivalent representation of the original delay system, together with the USP. Based on this representation, we give a controller parameterization and we solve the standard H 2 problem. Index Terms: dead-time compensator, modified Smith predictor, finite-spectrum assignment, finite-impulse-response (FIR) block, Youla parameterization, H 2 problem 1 Introduction Time delays appear in many physical systems, in particular, those involving material transportation and/or information transmission. Often, systems with delays appear as a simple approximation of more complex infinite-dimensional systems. The classical Smith predictor (SP) (Smith, Manuscript received on June 03, This research was supported by EPSRC under a Portfolio Partnership (Grant No. GR/S61256/01). Corresponding author, Tel: , Fax: , zhongqc@imperial.ac.uk, URL: 1
2 1957, 1958) is an effective tool to reduce control problems, such as pole assignment or tracking, for a finite-dimensional LTI stable system with an input or output delay, to corresponding delay-free problems. A finite spectrum assignment scheme was developed in (Olbrot, 1978; Manitius and Olbrot, 1979) to handle input delays in unstable plants by state feedback, using the predicted state of the delay system. Watanabe and Ito (1981) overcame some shortcomings in finite-spectrum assignment and other process-model-based control schemes available then, by using a Smith-predictor-like block, which was afterwards called a modified (or generalized) Smith predictor (MSP), see (Palmor, 1996) and the references therein. Recently, prediction has been recognized as a fundamental concept for the stabilization of delay systems (Mirkin and Raskin, 1999, 2003). Similar predictor blocks have played an important role in H control of time-delay systems (Meinsma and Zwart, 2000; Zhong, 2003b,c; Mirkin, 2003; Zhong, 2003e) and in continuous-time deadbeat control (Zhong, 2003a; Nobuyama et al., 1996). We show that the modified Smith predictor may run into numerical problems for delay systems with some fast stable eigenvalues. Indeed, the matrix exponential e Ah (where h is the delay) appearing in the MSP may be practically non-computable for such systems. Such a numerical problem was mentioned in (Meinsma and Zwart, 2000, p. 279) and a technique, which is not systematic, was suggested in (Zwart et al., 1998) to overcome the problem. This problem was attributed in these papers to large delays. Actually, this numerical problem might occur even for very small delays (with respect to the time constant of the system) if there are some stable eigenvalues λ and they are very fast with respect to the delay h (i.e., the product Reλ h is large). We propose an alternative predictor, called the unified Smith predictor (USP), to overcome this problem. The USP combines the features of the SP and the MSP and it does not require the computation of the matrix exponential for the stable eigenvalues. We achieve this by decomposing the state space of the finite-dimensional part of the plant into unstable and stable invariant subspaces. The controller design techniques based on the MSP have to be re-considered for the USP, to make them applicable in practice. For this reason, we propose an equivalent representation for the augmented plant, which consists of the original plant together with the USP. This equivalent representation is then used to give a stabilizing controller parameterization and to solve the standard H 2 problem. Further research is needed to solve the H control problem for a delay system with a USP. 2
3 This paper is organized as follows: The numerical problem with the MSP is explained in Section 2 using a very simple example, and then the USP is introduced. An equivalent representation of the augmented plant is derived in Section 3. We show in Section 4 how this equivalent representation can be used to derive a stabilizing controller parameterization and to solve the H 2 problem. 2 The unified Smith predictor This paper is written from a frequency domain perspective, which means that LTI systems are represented by their transfer functions. Here, by a transfer function we mean an analytic function defined on a domain which contains a half-plane C α = {s C Res >α}. A transfer function is called exponentially stable if it is bounded on a half-plane C α with α<0. Obviously, exponentially stable transfer functions are contained in H. We will use notation (common in the literature) P = A C B D, (1) meaning that P (s) =C(sI A) 1 B + D, where A, B, C, D are matrices. 2.1 The numerical problem with the MSP Consider a dead-time plant P h, with P h (s) =P (s)e sh, where h>0and the rational part P is realized as in (1). By a predictor for P h we mean an exponentially stable system Z such that P h + Z is rational. A predictor-based controller for the dead-time plant P h consists of a predictor Z and a stabilizing compensator C, as shown in Figure 1 (see also Figure 5 for a more general structure). The underlying idea is the well-known fact that there is a one-to-one correspondence between the set of all the stabilizing controllers for P h and for P h + Z, see for example Remark 3.6 and Example 4.1 in (Curtain et al., 1996). In order to simplify the exposition, only the stabilization problem will be studied in this section. If the tracking problem is considered, then an additional constraint on C is required, e.g., if the reference r is a step signal then it is required that lim s 0 C(s)(I+Z(s)C(s)) 1 =, see (Watanabe and Ito, 1981; Palmor, 1996) for more details. 3
4 If P is stable, then the predictor Z can be chosen to be the classical Smith predictor (SP) (Smith, 1957, 1958), Z SP (s) =P (s) P (s)e sh, (2) and the stabilizing compensator C is designed as a stabilizing controller for the delay-free system P. IfP is unstable, then the predictor Z can be chosen to be the modified Smith predictor (MSP) (Watanabe and Ito, 1981; Palmor, 1996), Z MSP (s) =P aug (s) P (s)e sh, (3) where P aug = P h + Z MSP, the augmented plant,isgivenby P aug = A B = A Ce Ah 0 C 0 e Ah B. (4) Note that Z MSP has its impulse response supported on [0, h], hence it is exponentially stable. Now the stabilizing compensator C is designed as a stabilizing controller for P aug. For further background and applications of the MSP we refer to (Meinsma and Zwart, 2000; Mirkin and Raskin, 2003). Implementing Z MSP is a delicate problem, because we have to avoid hidden unstable modes. One approach is to use some form of numerical integration using the past values of u (Palmor, 1996; Manitius and Olbrot, 1979; Zhong, 2003d). Alternatively, Z MSP can be approximately implemented by an FIR block with a stable rational part, so that it does not have hidden unstable modes (Watanabe and Ito, 1981). Recently, an interesting implementation of Z MSP using resetting integrators has been proposed in (Mondié et al., 2001). Now consider a simple example with P = , i.e., P (s) = According to (3), the predictor needed is s+1000 s 1 Z MSP (s) = e1000h e sh s e h e sh. s 1 Clearly, there is a numerical problem: e 1000h is a huge number even for a not so large delay h! Indeed, according to IEEE Standard 754 (IEEE, 1985), which is today the most common 4
5 r e C(s) u P (s)e sh y v Z(s) Figure 1: A dead-time plant with a predictor-based controller. The controller consists of the predictor Z (which is exponentially stable) and the stabilizing compensator C. Z is chosen such that the augmented plant P aug (s) =Z(s)+P (s)e sh is rational. Then C can be designed as a stabilizing controller for P aug. representation for real numbers on computers, including Intel-based PC s, Macintoshes and most Unix platforms, this number is regarded to be + (INF) for h 0.71sec. Certainly, such a component in a controller is not allowed in a practical application. We observe that this problem arises due to a fast stable pole (here, p = 1000) ofp. A stable pole makes the real part of the exponent positive, and if this is large, the numerical problem occurs. If the plant is completely unstable (there are no stable poles), then there will be no such problem. Thus, if the plant has fast stable poles, then the predictor Z MSP from (3) should not be used. An alternative predictor will be introduced below. 2.2 The unified Smith predictor A natural solution to the numerical problem encountered in the previous subsection is to decompose P into the sum of a stable part P s and an unstable part P u and then to construct predictors for them, using the classical Smith predictor (2) for P s (s)e sh and the modified Smith predictor (3) for P u (s)e sh. We propose a new term, unified Smith predictor (USP), for such a predictor. As is well known, the rational part of the plant, P given in (1), can be split (decomposed) into the sum of a stable part P s and an unstable part P u with ease, e.g., by applying a suitable linear coordinate transformation in its state space. There exist a lot of such similarity transformations. One of them, denoted here by V, can be obtained by bringing the system matrix A to the Jordan canonical form J = V 1 AV. (In MATLAB TM, this is done by [ V J ]=jordan(a). ) Assume 5
6 that V is a nonsingular matrix such that P = V 1 AV V 1 B CV D = A u 0 B u 0 A s B s C u C s D, (5) where A s is stable and A u is completely unstable. As a matter of fact, the decomposition can be done by splitting the complex plane along any vertical line Re s = α with α 0. Then the eigenvalues of A u are all the eigenvalues λ of A with Re λ α, while A s has the remaining eigenvalues of A. In the sequel, in order to simplify the exposition, we use the imaginary axis (α =0) to split the complex plane. Now P can be split as P = P s + P u, where B s P s = A s C s 0 and P u = A u The predictor for the stable part P s can be taken as a classical SP, C u B u D. Z s (s) =P s (s) P s (s)e sh, (6) and the predictor for the unstable part P u can be taken as the following MSP: Z u (s) =Pu aug (s) P u (s)e sh, Pu aug. A = u B u. C u e Auh 0 Then the USP for the plant P h (s) =P (s)e sh is defined by Z = Z s + Z u, as shown in Figure 2. It is now clear that Z(s) =P aug (s) P (s)e sh, (7) where P aug = P s + Pu aug and a realization of P aug is P aug = A B = A E hb, CE h 0 C 0 E h = V e Auh 0 0 I s V 1. (8) Here, the identity I s has the same dimension as A s and we have used that E h A = AE h. The impulse response of the USP is not finite, unlike for the MSP. The above P aug is the augmented plant, obtained by connecting the original plant and the USP in parallel. The stabilizing compensator C in Figure 1 should be designed as a stabilizing controller for P aug. 6
7 Pu aug (s) P u (s)e sh v Z u u P s (s) P s (s)e sh Z s Figure 2: The unified Smith predictor Z = Z s + Z u. Here, Z s is a classical SP for P s, the stable part of P, and Z u is an MSP for P u, the unstable part of P. The use of the different predictors (SP, MSP and USP) is summarized in Table 1: when the plant is stable, the USP reduces to the SP; when the plant is completely unstable, the USP reduces to the MSP; when the plant is unstable but has some stable poles (mixed), then the USP should be used, especially if some stable poles are fast (as explained earlier). Type of plant stable completely unstable mixed Type of USP SP MSP USP Table 1: The type of USP needed for different types of plants Returning to the example at the beginning of this section, the USP for it is given by Z(s) = 1 e sh s e h e sh. s 1 Remark 1. A possible realization of Z consists of the finite-dimensional system ż(t) =Az(t)+E h Bu(t) Bu(t h) v(t) =Cz(t) Du(t h) together with a realization of a delay line (which is needed to generate the signal u(t h)). However, such a realization of Z would have hidden unstable modes. An alternative is to realize Z s and Z u separately. The first is not a problem (it can be implemented dynamically), while for Z u see the comments after (4) and Remark 4. 7
8 Remark 2. Z s defined in (6) could be replaced by Z s (s) = P s (s)e sh, as in internal model control (Morari and Zafiriou, 1989). In this case, P aug in (8) would change, since in the definition of E h we would have to replace the identity I s by zero. Now P aug would be simpler (it would have a lower order). This may be advantageous in the context of Figure 1. However, in the more general framework of Figure 5 (in the next section) this may cause a rank problem (as we shall see). Remark 3. The USP (in particular, the SP and the MSP) can be generalized to multiple delays, i.e., to the situation where the components u 1, u 2,...u m of the vector u are delayed by h 1, h 2,...h m 0. Denote H = diag(h 1,h 2,... h m ) and [ E B = E h1 b 1 E h2 b 2... ] E hm b m, [ where b k is the k-th column of B, i.e., B = b 1 b 2... b m ]. Then and Z is exponentially stable. P aug = A E B, Z(s) =P aug (s) P (s)e Hs (9) C 0 Remark 4. We outline an implementation of Z from Remark 3 (the USP for multiple delays) which avoids the problem of unstable modes. We use the notation from (5). We realize the components Z u and Z s separately, but use the same delay lines for both, see Figure 3. The component Z u is a hybrid system, containing two copies of an LTI block with the possibility of resetting, and the switches S a, S b and S v. In this diagram, we have denoted [ ] [ B u = β 1 β 2... β m and E B u = e Auh 1 β 1 e Auh 2 β 2... e Auhm β m ]. The output of the USP is v = v s + v u, where v s = Z s u and v u = Z u u. v u is produced by one of the two LTI blocks which can be reset by the signals R a and R b, respectively. When we reset one of them, say the top one, at a time τ, we also switch its second input u a from u d to 0 using the switch S a. Afterwards, we reconnect u a k to ud k at the time τ + h k (k =1, 2,...m). During this time, v u = v b. After all the components of u d have been reconnected to the upper LTI block, so that u a = u d, its output v a will be the desired value of v u, and so the switch S v can be set so 8
9 that v u = v a. The switches could, theoretically, remain in this position forever, but since A u is unstable, tiny computational or rounding errors or the effect of noise will grow and will corrupt the output v a and hence v u. To prevent this, we reset the other LTI block (while switching the components of u b, similarly as above, using the switch S b ) and, when the output v b becomes correct, we switch S v so that v u = v b. This cycle repeats itself indefinitely and it can be shown that the USP is stable and produces the correct output. This implementation is related to the resetting Smith predictor of (Mondié et al., 2001). v v u v a S v v b A u E B u B u C u 0 D R a A u B u E B u C u D 0 R b u a S a 0 u b S b u d e Hs u v s A s B s B s C s 0 0 u d Z s Figure 3: An implementation of the USP (9) for multiple delays, using two resetting LTI systems to implement the component Z u. The logical controller which controls the switches and generates the resetting signals R a and R b (in an open loop manner) is not shown. 3 Control systems with a USP equivalent diagrams In this section we consider a more general type of plant, with two inputs and two outputs (each of these signals may be vector-valued). The input signal w contains references and disturbances, u is the control input, z is the tracking error and y is the measurement available to the controller. Such a plant is commonly considered in robust control, see for example (Green and Limebeer, 1995; Zhou et al., 1996). We assume that the plant P h consists of a rational part P and a delay 9
10 by h>0acting on u, as shown in Figure 4. We denote P = P A B 1 B 2 11 P 12 = C 1 D 11 D 12 P 21 P 22 C 2 D 21 D 22. (10) K is called a stabilizing controller for P h if the transfer functions from the three external inputs shown in Figure 4 to any other signal in the diagram are in H. K is called an exponentially stabilizing controller for P h if the same transfer functions are exponentially stable. In the sequel, the two external signals appearing in the lower part of Figure 4 will be taken to be zero. z y P h (s) P (s) e sh w u K(s) Figure 4: The control system comprising a dead-time plant P h (with a rational part P ) and a stabilizing controller K. While P may be unstable, we would like the control system to be (exponentially) stable and the transfer function from w to z to be small. As in the previous section, let V be a nonsingular matrix such that V 1 AV = A u 0, 0 A s where A u is completely unstable and A s is stable. As in (7), we introduce the following USP designed for the component P 22 of the plant: Z(s). = P aug 22 (s) P 22 (s)e sh, (11) where, as in (8), P aug 22. = A B 2 C 2 E h 0, E h = V e Auh 0 0 I s V 1. 10
11 By connecting this USP in parallel with the u to y component of P h, as shown in Figure 5, thus creating a new measurement output y p, we obtain a new augmented plant P aug (s) = P 11(s) P 21 (s) P 12 (s)e hs P aug 22 (s). (12) It is well known (and easy to see from Figure 5) that C is a stabilizing controller for P aug if and only if K = C(I ZC) 1 is a stabilizing controller for the original dead-time plant. The same statement remains true with exponentially stabilizing in place of stabilizing (and, of course, the corresponding set of controllers is smaller). z y P (s) e sh w Z(s) y p P aug (s) u C(s) Figure 5: The control system from Figure 4, in which the controller K has been decomposed into a USP denoted Z and a stabilizing compensator C, so that K = C(I ZC) 1. The augmented plant P aug consists of the plant P h together with the USP. We derive now an equivalent representation of P aug, which is useful for the problems treated in the next section. Proposition 1 and Remark 5 below are related to Lemma 1 in (Mirkin and Raskin, 2003) (the corresponding notation is 1 = Z 1 and 2 = Z). Proposition 1. We have P aug (s) = Z 1(s) e sh I 0 0 I P (s), (13) where Z 1 (s). = P 11 (s) A eah B 1 C 1 0 (s)e sh, 11
12 P = P 0 + P A E 1 h s 0 B 1 B 2, P0 = C 1 0 D , (14) C 2 E h D 21 0 [ ] A s 0 e Ash I s V 1 B 1 P s = C 1 V 0 0. (15) I s The block diagram corresponding to the decomposition (13)-(14) of P aug is in Figure 6. Z 1 (s) z z P s (s) e sh w P z 0 P 0 (s) y p C(s) u Figure 6: An equivalent representation of the control system in Figure 5, using the decomposition of P aug given in Proposition 1. Here, Z 1 and P s are exponentially stable. Proof. Using the partition (12) of P aug, the formula (13) (which we have to prove) is equivalent to the following four formulas: P 11 (s) =Z 1 (s)+ A C 1 0 P 21 = E 1 h B 1 P 12 (s)e sh = A B 2 (s)e sh, C 1 D 12 A E 1 h B 1 C 2 E h D 21, (s)+ P s (s) e sh, (16) P aug 22 = A B 2 C 2 E h 0. 12
13 The last three formulas are clearly true, so only the first formula (16) requires a little work. If we rewrite A E 1 h B 1 C 1 0 = A u 0 eauh 0 0 A s 0 I s C 1 V 0 V 1 B 1 and if we rewrite P s = A u 0 0 A s 0 0 V 1 B 1 0 e Ash I s C 1 V 0, then (16) follows easily. Note that if we would choose the USP described in Remark 2, then E h would become singular and hence the decomposition (14) of P would have to be replaced by a more complicated one. This is the rank problem mentioned in Remark 2. Remark 5. It follows from Proposition 1 that the closed-loop transfer function T zw from w to z can be written as the sum of three terms: T zw (s) =Z 1 (s)+ P s (s)e sh + F l ( P 0 (s),c(s))e sh, (17) where F l ( P 0,C) is the transfer function from w to z 0 in Figure 6. Z 1 is an FIR system with impulse response supported on [0,h], while the second and the third term have impulse responses supported on [h, ). Hence, if these terms are in H 2 (C 0 ), then the first term is orthogonal to the second and third term. Remark 6. A realization of P is given by A 0 E 1 h B 1 B 2 [ ] 0 A s 0 e Ash I s V 1 B 1 0 P = C 1 C 1 V 0. (18) 0 D 12 I s C 2 E h 0 D 21 0 If P is completely unstable, then the dimensions of A s and I s are zero and the block P s in 13
14 Figure 6 disappears. In this case, Z reduces to an MSP and P becomes A e Ah B 1 B 2 P = C 1 0 D 12 = P 0. C 2 e Ah D 21 0 If P is stable, then P is reduced to 4 Applications P = A 0 B 1 B 2 0 A (e Ah I)B 1 0 C 1 C 1 0 D 12 C 2 0 D The parameterization of all stabilizing controllers We consider the control system in Figure 4 and we derive necessary and sufficient conditions on P h for the existence of an exponentially stabilizing controller K. It is known that in general, exponential stabilization for an infinite-dimensional plant is more difficult to achieve than stabilization (in the H sense) (Weiss et al., 2001). It turns out that in our particular setting, the two problems are equivalent, and can be reduced to the stabilization of P 0 from (14). Theorem 2. The dead-time plant P h shown in Figure 4 with a minimal realization of its rational part as in (10) admits an (exponentially) stabilizing controller K (as defined in Section 3) if and only if (A, B 2 ) is stabilizable and (C 2,A) is detectable. With the USP Z given in (11), every such controller can be expressed as K = C(I ZC) 1, (19) where C is an (exponentially) stabilizing controller for P 0 defined in (14). Let F and L be such that A + LC 2 and A + B 2 F are stable, then every stabilizing C for P 0 can be expressed as A + B 2 F + E 1 h LC 2E h E 1 h L B 2 C = F l (M,Q), M = F 0 I (20) C 2 E h I 0 14
15 where Q H. Such a C is exponentially stabilizing if and only if Q is exponentially stable. The closed-loop transfer function T zw achieved by K from (19) is given by (17), where N = F l ( P 0,C)=F l (N,Q), A + B 2 F B 2 F E 1 h B 1 B 2 0 A + E 1 h LC 2E h E 1 h (B 1 + LD 21 ) 0 C 1 + D 12 F D 12 F 0 D 12 0 C 2 E h D Proof. As explained in the previous section, the control system in Figure 4 has an equivalent representation shown in Figure 6. The blocks Z 1 and P s in Figure 6 are exponentially stable, so that they have no influence on the (exponential) stability of the whole system (since they are not a part of any feedback loop). Thus, the (exponential) stability of the original closed-loop system is equivalent to that of the system formed from P 0 and C only. First we discuss stabilization in the H sense. Since P 0 is rational, the parameterization of the stabilizing controllers using M follows from (Zhou et al., 1996, p. 312) where we replace C 2 by C 2 E h and L by E 1 h L. This is possible because A + LC 2 is similar to A + E 1 h LC 2E h (hence, A + LC 2 is stable if and only if A + E 1 h LC 2E h is stable). For a rigorous derivation of the parameterization for irrational plants, see (Curtain et al., 2001). The formula involving N follows from (Zhou et al., 1996, p. 323). Now we consider the exponential stabilization. Let α<0 be such that all the eigenvalues of A + B 2 F and A + LC 2 are in the half-plane where Re s<α. Now all the arguments used to derive the parameterization (20) of stabilizing controllers can be redone with H replaced by H (C α ), the space of bounded analytic functions on C α. This will result in the same formulae (20), but now Q must be in H (C α ). 4.2 The H 2 problem Consider again the feedback system from Figure 4. The H 2 problem is to find a stabilizing controller K which minimizes the H 2 -norm of the transfer function T zw. The solution of this problem, using the modified Smith predictor, is well known, see (Mirkin and Raskin, 1999, 15
16 2003) and the references therein. In this section, we will re-consider this problem using the USP. We shall work with a minimal realization of P, of the form (10). Assume the following: (A1) (A, B 2 ) is stabilizable and (C 2,A) is detectable; (A2) A jωi B 2 has full column rank ω R; C 1 D 12 (A3) A jωi B 1 has full row rank ω R; C 2 D 21 (A4) D12D 12 = I and D 21 D21 = I. Assumption (A4) is made just to simplify the exposition. In fact, only the non-singularity of D12 D 12 and D 21 D21 is required (Green and Limebeer, 1995; Zhou and Doyle, 1998). As mentioned in Remark 5, the impulse response of Z 1 is supported on [0,h] and the last two terms of T zw in (17) (which together are F l ( P,C) delayed by h) are supported on [h, ). Assume for a moment that D 11 =0. Then it follows that Z 1 is orthogonal (in H 2 ) to the last two terms of T zw, so that T zw 2 2 = Z Fl ( P,C)e sh 2 2 = Z Fl ( P,C). The H 2 control problem of minimizing T zw 2 over all stabilizing K is then converted to γ =min F l ( P,C) (C stabilizing). C 2 This is a finite-dimensional H 2 problem which can be solved using known results, see e.g. (Zhou et al., 1996, Theorem 14.7 on p. 385). This last problem is meaningful even if D 11 0: we are then minimizing the L 2 norm of the impulse response of T zw restricted to [h, ). Note that the restriction to [0, h] is independent of C. Rewrite P from (18) as à B1 B2 P = C 1 0 D 12, C 2 D 21 0 then the solution of the H 2 problem involves the following two Hamiltonian matrices: à 0 B H 2 = C 1 C 2 [ 1 à C 1 D D C ] 12 1 B 2,
17 Ã J 2 = 0 C 2 B [ ] 1 B 1 Ã B D 21 B 1 C2. 1 D21 When the conditions (A1-A4) hold, these two Hamiltonian matrices belong to dom(ric), as defined in (Zhou et al., 1996), and moreover, X 2. = Ric(H2 ) 0 and Y 2. = Ric(J2 ) 0. Theorem 3. There exists a unique optimal H 2 controller K = C(I ZC) 1, C = Ã + B 2 F 2 + L 2 C2 L 2 F 2 0 where, F 2 = ( B 2X 2 + D12 C 1 ), L 2 = (Y 2 C 2 + B 1 D21). Proof. The only thing to be verified is whether the conditions (A1)-(A4) hold for P. Using the zeros in the matrix in (18), we can see that this is true, since A s is stable. The solution can be obtained by applying Theorem 14.7 from (Zhou et al., 1996, p. 385). 5 Discussions and conclusions We have pointed out a numerical problem with the modified Smith predictor when the plant has fast stable poles and we have introduced the unified Smith predictor as a remedy. We have derived an equivalent representation of the augmented plant consisting of a dead-time plant and a unified Smith predictor. Using this representation, we have derived a parameterization of the (exponentially) stabilizing controllers for the dead-time plant (with the USP connected to it) and we have solved the standard H 2 problem (again, in the presence of the USP). The same numerical problem may arise also in the solution of the H control problem for dead-time plants (Meinsma and Zwart, 2000; Zhong, 2003b; Mirkin, 2003; Zhong, 2003e), where the controller design involves the computation of e Ah. A reasonable assumption is that the computation of e Ah does not cause such a problem (due to the unstable modes in A), because otherwise the system blows up before the control can have any effect. Acknowledgment The idea of building a predictor in two parts, corresponding to the stable and the unstable part of the plant, arose out of discussions with L. Mirkin. 17
18 References Curtain, R., G. Weiss and M. Weiss (1996). Coprime factorization for regular linear systems. Automatica 32(11), Curtain, R., G. Weiss and M. Weiss (2001). Stabilization of irrational transfer functions by controllers with internal loop. In: Operator Theory: Advances and Applications (A.A. Borichev and N.K. Nikolskü, Eds.). Vol pp Birkhäuser. Green, M. and D.J.N. Limebeer (1995). Linear Robust Control. Prentice-Hall. Englewood Cliffs, NJ. IEEE (1985). ANSI/IEEE , Standard for Binary Floating-Point Arithmetic. IEEE. New York, NY, USA. Manitius, A.Z. and A.W. Olbrot (1979). Finite spectrum assignment problem for systems with delays. IEEE Trans. Automat. Control 24(4), Meinsma, G. and H. Zwart (2000). On H control for dead-time systems. IEEE Trans. Automat. Control 45(2), Mirkin, L. (2003). On the extraction of dead-time controllers and estimators from delay-free parametrizations. IEEE Trans. Automat. Control 48(4), Mirkin, L. and N. Raskin (1999). Every stabilizing dead-time controller has an observer-predictor-based structure. In: Proc. of the 7th IEEE Mediterranean Conference in Control and Automation. Haifa, Israel. pp Mirkin, L. and N. Raskin (2003). Every stabilizing dead-time controller has an observer-predictor-based structure. Automatica. to appear. Mondié, S., R. Lozano and J. Collado (2001). Resetting process-model control for unstable systems with delay. In: Proc. of the 40th IEEE Conference on Decision & control. Vol. 3. Orlando, Florida, USA. pp Morari, M. and E. Zafiriou (1989). Robust Process Control. Prentice-Hall, Inc. Nobuyama, E., S. Shin and T. Kitamori (1996). Deadbeat control of continous-time systems: MIMO case. In: Proc. of the 35th IEEE Conference on Decision & control. Kobe, Japan. pp
19 Olbrot, A.W. (1978). Stabilizability, detectability, and spectrum assignment for linear autonomous systems with general time delays. IEEE Trans. Automat. Control 23(5), Palmor, Z.J. (1996). Time-delay compensation Smith predictor and its modifications. In: The Control Handbook (S. Levine, Ed.). pp CRC Press. Smith, O.J.M. (1957). Closer control of loops with dead time. Chem. Eng. Progress 53(5), Smith, O.J.M. (1958). Feedback Control Systems. McGraw-Hill Book Company, Inc. Watanabe, K. and M. Ito (1981). A process-model control for linear systems with delay. IEEE Trans. Automat. Control 26(6), Weiss, G., O.J. Staffans and M. Tucsnak (2001). Well-posed linear systems a survey with emphasis on conservative systems. Int. J. Appl. Math. Comput. Sci. 11(1), Zhong, Q.-C. (2003a). Control of integral processes with dead-time. Part 3: Deadbeat disturbance response. IEEE Trans. Automat. Control 48(1), Zhong, Q.-C. (2003b). Frequency domain solution to delay-type Nehari problem. Automatica 39(3), Zhong, Q.-C. (2003c). H control of dead-time systems based on a transformation. Automatica 39(2), Zhong, Q.-C. (2003d). The holding effect hidden in quadrature approximations: Implementation of distributed delays in control laws. Submitted to IEEE Trans. AC. Zhong, Q.-C. (2003e). On standard H control of processes with a single delay. IEEE Trans. Automat. Control 48(6),. Zhou, K. and J.C. Doyle (1998). Essentials of Robust Control. Prentice-Hall. Upper Saddle River, N.J. Zhou, K., J.C. Doyle and K. Glover (1996). Robust and Optimal Control. Prentice-Hall. Englewood Cliffs, NJ. Zwart, H., G. Weiss and G. Meinsma (1998). Prediction of a narrow-band signal from measurement data. In: Proc. of the 4th International Conference on Optimization: Techniques and Applications. Perth, Australia. pp
A Design Method for Smith Predictors for Minimum-Phase Time-Delay Plants
00 ECTI TRANSACTIONS ON COMPUTER AND INFORMATION TECHNOLOGY VOL., NO.2 NOVEMBER 2005 A Design Method for Smith Predictors for Minimum-Phase Time-Delay Plants Kou Yamada Nobuaki Matsushima, Non-members
More informationA Case Study for the Delay-type Nehari Problem
Proceedings of the 44th EEE Conference on Decision and Control, and the European Control Conference 5 Seville, Spain, December -5, 5 MoA A Case Study for the Delay-type Nehari Problem Qing-Chang Zhong
More informationRobust Control of Time-delay Systems
Robust Control of Time-delay Systems Qing-Chang Zhong Distinguished Lecturer, IEEE Power Electronics Society Max McGraw Endowed Chair Professor in Energy and Power Engineering Dept. of Electrical and Computer
More informationJ-SPECTRAL FACTORIZATION
J-SPECTRAL FACTORIZATION of Regular Para-Hermitian Transfer Matrices Qing-Chang Zhong zhongqc@ieee.org School of Electronics University of Glamorgan United Kingdom Outline Notations and definitions Regular
More informationLyapunov Stability of Linear Predictor Feedback for Distributed Input Delays
IEEE TRANSACTIONS ON AUTOMATIC CONTROL VOL. 56 NO. 3 MARCH 2011 655 Lyapunov Stability of Linear Predictor Feedback for Distributed Input Delays Nikolaos Bekiaris-Liberis Miroslav Krstic In this case system
More informationThe parameterization of all. of all two-degree-of-freedom strongly stabilizing controllers
The parameterization stabilizing controllers 89 The parameterization of all two-degree-of-freedom strongly stabilizing controllers Tatsuya Hoshikawa, Kou Yamada 2, Yuko Tatsumi 3, Non-members ABSTRACT
More informationTHIS paper deals with robust control in the setup associated
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL 50, NO 10, OCTOBER 2005 1501 Control-Oriented Model Validation and Errors Quantification in the `1 Setup V F Sokolov Abstract A priori information required for
More informationClosed-Loop Structure of Discrete Time H Controller
Closed-Loop Structure of Discrete Time H Controller Waree Kongprawechnon 1,Shun Ushida 2, Hidenori Kimura 2 Abstract This paper is concerned with the investigation of the closed-loop structure of a discrete
More informationThe Solvability Conditions for the Inverse Eigenvalue Problem of Hermitian and Generalized Skew-Hamiltonian Matrices and Its Approximation
The Solvability Conditions for the Inverse Eigenvalue Problem of Hermitian and Generalized Skew-Hamiltonian Matrices and Its Approximation Zheng-jian Bai Abstract In this paper, we first consider the inverse
More informationFEL3210 Multivariable Feedback Control
FEL3210 Multivariable Feedback Control Lecture 8: Youla parametrization, LMIs, Model Reduction and Summary [Ch. 11-12] Elling W. Jacobsen, Automatic Control Lab, KTH Lecture 8: Youla, LMIs, Model Reduction
More informationProjection of state space realizations
Chapter 1 Projection of state space realizations Antoine Vandendorpe and Paul Van Dooren Department of Mathematical Engineering Université catholique de Louvain B-1348 Louvain-la-Neuve Belgium 1.0.1 Description
More informationKrylov Techniques for Model Reduction of Second-Order Systems
Krylov Techniques for Model Reduction of Second-Order Systems A Vandendorpe and P Van Dooren February 4, 2004 Abstract The purpose of this paper is to present a Krylov technique for model reduction of
More informationGlobally Stabilizing Adaptive Control Design for Nonlinearly-Parameterized Systems REFERENCES
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 5, NO. 6, JUNE 26 73 an indicator of performance deterioration. Yet in general this might not be true. For example, consider the problem of the minimization
More informationModel reduction of interconnected systems
Model reduction of interconnected systems A Vandendorpe and P Van Dooren 1 Introduction Large scale linear systems are often composed of subsystems that interconnect to each other Instead of reducing the
More informationOn Positive Real Lemma for Non-minimal Realization Systems
Proceedings of the 17th World Congress The International Federation of Automatic Control On Positive Real Lemma for Non-minimal Realization Systems Sadaaki Kunimatsu Kim Sang-Hoon Takao Fujii Mitsuaki
More informationCANONICAL LOSSLESS STATE-SPACE SYSTEMS: STAIRCASE FORMS AND THE SCHUR ALGORITHM
CANONICAL LOSSLESS STATE-SPACE SYSTEMS: STAIRCASE FORMS AND THE SCHUR ALGORITHM Ralf L.M. Peeters Bernard Hanzon Martine Olivi Dept. Mathematics, Universiteit Maastricht, P.O. Box 616, 6200 MD Maastricht,
More information2DOF Controller Parametrization for Systems With a Single I/O Delay
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 48, NO. 11, NOVEMBER 2003 1999 For 62 6 o1 2 6 o2 M ((x 1 ;x 2 ;x 3 ;x 4 );) H (x 1;);x 2;x 3;x 4 x 1 ; H (x 2 ;); H (x 3 ;); G (x 4 ;) H (x 1 ;); H (x 2 ;);
More information6.241 Dynamic Systems and Control
6.241 Dynamic Systems and Control Lecture 24: H2 Synthesis Emilio Frazzoli Aeronautics and Astronautics Massachusetts Institute of Technology May 4, 2011 E. Frazzoli (MIT) Lecture 24: H 2 Synthesis May
More informationAn LQ R weight selection approach to the discrete generalized H 2 control problem
INT. J. CONTROL, 1998, VOL. 71, NO. 1, 93± 11 An LQ R weight selection approach to the discrete generalized H 2 control problem D. A. WILSON², M. A. NEKOUI² and G. D. HALIKIAS² It is known that a generalized
More informationIMC based automatic tuning method for PID controllers in a Smith predictor configuration
Computers and Chemical Engineering 28 (2004) 281 290 IMC based automatic tuning method for PID controllers in a Smith predictor configuration Ibrahim Kaya Department of Electrical and Electronics Engineering,
More informationFrequency domainsolutionto delay-type Nehari problem
Available online at www.sciencedirect.com Automatica 39 (2003) 499 508 www.elsevier.com/locate/automatica Brief paper Frequency domainsolutionto delay-type Nehari problem Qing-Chang Zhong Department of
More informationIMPULSIVE CONTROL OF DISCRETE-TIME NETWORKED SYSTEMS WITH COMMUNICATION DELAYS. Shumei Mu, Tianguang Chu, and Long Wang
IMPULSIVE CONTROL OF DISCRETE-TIME NETWORKED SYSTEMS WITH COMMUNICATION DELAYS Shumei Mu Tianguang Chu and Long Wang Intelligent Control Laboratory Center for Systems and Control Department of Mechanics
More informationDiscussion on: Measurable signal decoupling with dynamic feedforward compensation and unknown-input observation for systems with direct feedthrough
Discussion on: Measurable signal decoupling with dynamic feedforward compensation and unknown-input observation for systems with direct feedthrough H.L. Trentelman 1 The geometric approach In the last
More informationA Simple Derivation of Right Interactor for Tall Transfer Function Matrices and its Application to Inner-Outer Factorization Continuous-Time Case
A Simple Derivation of Right Interactor for Tall Transfer Function Matrices and its Application to Inner-Outer Factorization Continuous-Time Case ATARU KASE Osaka Institute of Technology Department of
More informationDynamic Model Predictive Control
Dynamic Model Predictive Control Karl Mårtensson, Andreas Wernrud, Department of Automatic Control, Faculty of Engineering, Lund University, Box 118, SE 221 Lund, Sweden. E-mail: {karl, andreas}@control.lth.se
More informationExponential stabilization of a Rayleigh beam - actuator and feedback design
Exponential stabilization of a Rayleigh beam - actuator and feedback design George WEISS Department of Electrical and Electronic Engineering Imperial College London London SW7 AZ, UK G.Weiss@imperial.ac.uk
More informationThe Rationale for Second Level Adaptation
The Rationale for Second Level Adaptation Kumpati S. Narendra, Yu Wang and Wei Chen Center for Systems Science, Yale University arxiv:1510.04989v1 [cs.sy] 16 Oct 2015 Abstract Recently, a new approach
More informationStructured State Space Realizations for SLS Distributed Controllers
Structured State Space Realizations for SLS Distributed Controllers James Anderson and Nikolai Matni Abstract In recent work the system level synthesis (SLS) paradigm has been shown to provide a truly
More informationTHE PARAMETERIZATION OF ALL ROBUST STABILIZING MULTI-PERIOD REPETITIVE CONTROLLERS FOR MIMO TD PLANTS WITH THE SPECIFIED INPUT-OUTPUT CHARACTERISTIC
International Journal of Innovative Computing, Information Control ICIC International c 218 ISSN 1349-4198 Volume 14, Number 2, April 218 pp. 387 43 THE PARAMETERIZATION OF ALL ROBUST STABILIZING MULTI-PERIOD
More informationLinear Quadratic Gausssian Control Design with Loop Transfer Recovery
Linear Quadratic Gausssian Control Design with Loop Transfer Recovery Leonid Freidovich Department of Mathematics Michigan State University MI 48824, USA e-mail:leonid@math.msu.edu http://www.math.msu.edu/
More informationMRAGPC Control of MIMO Processes with Input Constraints and Disturbance
Proceedings of the World Congress on Engineering and Computer Science 9 Vol II WCECS 9, October -, 9, San Francisco, USA MRAGPC Control of MIMO Processes with Input Constraints and Disturbance A. S. Osunleke,
More informationEEE582 Homework Problems
EEE582 Homework Problems HW. Write a state-space realization of the linearized model for the cruise control system around speeds v = 4 (Section.3, http://tsakalis.faculty.asu.edu/notes/models.pdf). Use
More informationSYSTEMTEORI - ÖVNING 5: FEEDBACK, POLE ASSIGNMENT AND OBSERVER
SYSTEMTEORI - ÖVNING 5: FEEDBACK, POLE ASSIGNMENT AND OBSERVER Exercise 54 Consider the system: ẍ aẋ bx u where u is the input and x the output signal (a): Determine a state space realization (b): Is the
More informationOptimal Polynomial Control for Discrete-Time Systems
1 Optimal Polynomial Control for Discrete-Time Systems Prof Guy Beale Electrical and Computer Engineering Department George Mason University Fairfax, Virginia Correspondence concerning this paper should
More informationRELAY CONTROL WITH PARALLEL COMPENSATOR FOR NONMINIMUM PHASE PLANTS. Ryszard Gessing
RELAY CONTROL WITH PARALLEL COMPENSATOR FOR NONMINIMUM PHASE PLANTS Ryszard Gessing Politechnika Śl aska Instytut Automatyki, ul. Akademicka 16, 44-101 Gliwice, Poland, fax: +4832 372127, email: gessing@ia.gliwice.edu.pl
More informationDESIGN OF ROBUST CONTROL SYSTEM FOR THE PMS MOTOR
Journal of ELECTRICAL ENGINEERING, VOL 58, NO 6, 2007, 326 333 DESIGN OF ROBUST CONTROL SYSTEM FOR THE PMS MOTOR Ahmed Azaiz Youcef Ramdani Abdelkader Meroufel The field orientation control (FOC) consists
More informationA New Subspace Identification Method for Open and Closed Loop Data
A New Subspace Identification Method for Open and Closed Loop Data Magnus Jansson July 2005 IR S3 SB 0524 IFAC World Congress 2005 ROYAL INSTITUTE OF TECHNOLOGY Department of Signals, Sensors & Systems
More informationOptimal triangular approximation for linear stable multivariable systems
Proceedings of the 007 American Control Conference Marriott Marquis Hotel at Times Square New York City, USA, July -3, 007 Optimal triangular approximation for linear stable multivariable systems Diego
More informationPositive Stabilization of Infinite-Dimensional Linear Systems
Positive Stabilization of Infinite-Dimensional Linear Systems Joseph Winkin Namur Center of Complex Systems (NaXys) and Department of Mathematics, University of Namur, Belgium Joint work with Bouchra Abouzaid
More informationObserver design for rotating shafts excited by unbalances
Observer design for rotating shafts excited by unbalances R. S. Schittenhelm, Z. Wang, S. Rinderknecht Institute for Mechatronic Systems in Mechanical Engineering, Technische Universität Darmstadt, Germany
More information554 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 55, NO. 2, FEBRUARY and such that /$ IEEE
554 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 55, NO. 2, FEBRUARY 2010 REFERENCES [1] M. Fliess, J. Levine, P. Martin, and P. Rouchon, Flatness and defect of nonlinear systems: Introductory theory and
More information3 Stabilization of MIMO Feedback Systems
3 Stabilization of MIMO Feedback Systems 3.1 Notation The sets R and S are as before. We will use the notation M (R) to denote the set of matrices with elements in R. The dimensions are not explicitly
More informationControl of integral processes with dead time Part IV: various issues about PI controllers
Control of integral processes with dead time Part IV: various issues about PI controllers B. Wang, D. Rees and Q.-C. Zhong Abstract: Various issues about integral processes with dead time controlled by
More informationMASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Electrical Engineering and Computer Science : Dynamic Systems Spring 2011
MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Electrical Engineering and Computer Science 6.4: Dynamic Systems Spring Homework Solutions Exercise 3. a) We are given the single input LTI system: [
More informationContents. 1 State-Space Linear Systems 5. 2 Linearization Causality, Time Invariance, and Linearity 31
Contents Preamble xiii Linear Systems I Basic Concepts 1 I System Representation 3 1 State-Space Linear Systems 5 1.1 State-Space Linear Systems 5 1.2 Block Diagrams 7 1.3 Exercises 11 2 Linearization
More informationROBUST STABILITY ANALYSIS OF SIMPLE CONTROL ALGORITHMS IN COMMUNICATION NETWORKS
ROBUST STABILITY ANALYSIS OF SIMPLE CONTROL ALGORITHMS IN COMMUNICATION NETWORKS Qing-Chang Zhong Dept of Electrical & Electronic Engineering Imperial College London Exhibition Rd, London SW7 BT United
More informationIs Monopoli s Model Reference Adaptive Controller Correct?
Is Monopoli s Model Reference Adaptive Controller Correct? A. S. Morse Center for Computational Vision and Control Department of Electrical Engineering Yale University, New Haven, CT 06520 USA April 9,
More informationPreface. Figures Figures appearing in the text were prepared using MATLAB R. For product information, please contact:
Linear algebra forms the basis for much of modern mathematics theoretical, applied, and computational. The purpose of this book is to provide a broad and solid foundation for the study of advanced mathematics.
More informationA Convex Characterization of Distributed Control Problems in Spatially Invariant Systems with Communication Constraints
A Convex Characterization of Distributed Control Problems in Spatially Invariant Systems with Communication Constraints Bassam Bamieh Petros G. Voulgaris Revised Dec, 23 Abstract In this paper we consider
More informationResearch Article Stabilization Analysis and Synthesis of Discrete-Time Descriptor Markov Jump Systems with Partially Unknown Transition Probabilities
Research Journal of Applied Sciences, Engineering and Technology 7(4): 728-734, 214 DOI:1.1926/rjaset.7.39 ISSN: 24-7459; e-issn: 24-7467 214 Maxwell Scientific Publication Corp. Submitted: February 25,
More informationStatic Output Feedback Stabilisation with H Performance for a Class of Plants
Static Output Feedback Stabilisation with H Performance for a Class of Plants E. Prempain and I. Postlethwaite Control and Instrumentation Research, Department of Engineering, University of Leicester,
More informationTHE WINDSURFER APPROACH TO ITERATIVE CONTROL AND IDENTIFICATION USING AN H ALGORITHM. Arvin Dehghani,,1 Alexander Lanzon, Brian D. O.
IFA Workshop on Adaptation and Learning in ontrol and Signal rocessing, and IFA Workshop on eriodic ontrol Systems, Yokohama, Japan, August 30 September 1, 2004 THE WINDSURFER AROAH TO ITERATIVE ONTROL
More informationControl Systems. Laplace domain analysis
Control Systems Laplace domain analysis L. Lanari outline introduce the Laplace unilateral transform define its properties show its advantages in turning ODEs to algebraic equations define an Input/Output
More informationModel reduction via tangential interpolation
Model reduction via tangential interpolation K. Gallivan, A. Vandendorpe and P. Van Dooren May 14, 2002 1 Introduction Although most of the theory presented in this paper holds for both continuous-time
More informationLMI Based Model Order Reduction Considering the Minimum Phase Characteristic of the System
LMI Based Model Order Reduction Considering the Minimum Phase Characteristic of the System Gholamreza Khademi, Haniyeh Mohammadi, and Maryam Dehghani School of Electrical and Computer Engineering Shiraz
More informationMassachusetts Institute of Technology. Department of Electrical Engineering and Computer Science : MULTIVARIABLE CONTROL SYSTEMS by A.
Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science 6.245: MULTIVARIABLE CONTROL SYSTEMS by A. Megretski Q-Parameterization 1 This lecture introduces the so-called
More informationRational Implementation of Distributed Delay Using Extended Bilinear Transformations
Rational Implementation of Distributed Delay Using Extended Bilinear Transformations Qing-Chang Zhong zhongqc@ieee.org, http://come.to/zhongqc School of Electronics University of Glamorgan United Kingdom
More informationParametrization of All Strictly Causal Stabilizing Controllers of Multidimensional Systems single-input single-output case
Parametrization of All Strictly Causal Stabilizing Controllers of Multidimensional Systems single-input single-output case K. Mori Abstract We give a parametrization of all strictly causal stabilizing
More informationOptimal Controller Initialization for Switching between Stabilizing Controllers
Optimal Controller Initialization for Switching between Stabilizing Controllers João P. Hespanha, Paolo Santesso, Gregory Stewart Abstract An important class of switched systems consists of those systems
More informationFinding eigenvalues for matrices acting on subspaces
Finding eigenvalues for matrices acting on subspaces Jakeniah Christiansen Department of Mathematics and Statistics Calvin College Grand Rapids, MI 49546 Faculty advisor: Prof Todd Kapitula Department
More informationAn Architecture for Fault Tolerant Controllers
An Architecture for Fault Tolerant Controllers Henrik Niemann Ørsted DTU, Automation Technical University of Denmark Building 326, DK2800 Lyngby, Denmark Email: hhn@oersted.dtu.dk www.oersted.dtu.dk/staff/all/hhn/
More informationZeros and zero dynamics
CHAPTER 4 Zeros and zero dynamics 41 Zero dynamics for SISO systems Consider a linear system defined by a strictly proper scalar transfer function that does not have any common zero and pole: g(s) =α p(s)
More informationObserver Based Friction Cancellation in Mechanical Systems
2014 14th International Conference on Control, Automation and Systems (ICCAS 2014) Oct. 22 25, 2014 in KINTEX, Gyeonggi-do, Korea Observer Based Friction Cancellation in Mechanical Systems Caner Odabaş
More informationInternal Model Control of A Class of Continuous Linear Underactuated Systems
Internal Model Control of A Class of Continuous Linear Underactuated Systems Asma Mezzi Tunis El Manar University, Automatic Control Research Laboratory, LA.R.A, National Engineering School of Tunis (ENIT),
More informationStability, Pole Placement, Observers and Stabilization
Stability, Pole Placement, Observers and Stabilization 1 1, The Netherlands DISC Course Mathematical Models of Systems Outline 1 Stability of autonomous systems 2 The pole placement problem 3 Stabilization
More informationRECURSIVE SUBSPACE IDENTIFICATION IN THE LEAST SQUARES FRAMEWORK
RECURSIVE SUBSPACE IDENTIFICATION IN THE LEAST SQUARES FRAMEWORK TRNKA PAVEL AND HAVLENA VLADIMÍR Dept of Control Engineering, Czech Technical University, Technická 2, 166 27 Praha, Czech Republic mail:
More informationMultivariable ARMA Systems Making a Polynomial Matrix Proper
Technical Report TR2009/240 Multivariable ARMA Systems Making a Polynomial Matrix Proper Andrew P. Papliński Clayton School of Information Technology Monash University, Clayton 3800, Australia Andrew.Paplinski@infotech.monash.edu.au
More informationGlobal Stabilization of Oscillators With Bounded Delayed Input
Global Stabilization of Oscillators With Bounded Delayed Input Frédéric Mazenc, Sabine Mondié and Silviu-Iulian Niculescu Abstract. The problem of globally asymptotically stabilizing by bounded feedback
More informationUnbiased minimum variance estimation for systems with unknown exogenous inputs
Unbiased minimum variance estimation for systems with unknown exogenous inputs Mohamed Darouach, Michel Zasadzinski To cite this version: Mohamed Darouach, Michel Zasadzinski. Unbiased minimum variance
More informationControl Systems Design
ELEC4410 Control Systems Design Lecture 14: Controllability Julio H. Braslavsky julio@ee.newcastle.edu.au School of Electrical Engineering and Computer Science Lecture 14: Controllability p.1/23 Outline
More informationStabilization of Distributed Parameter Systems by State Feedback with Positivity Constraints
Stabilization of Distributed Parameter Systems by State Feedback with Positivity Constraints Joseph Winkin Namur Center of Complex Systems (naxys) and Dept. of Mathematics, University of Namur, Belgium
More informationFeedback Stabilization over Signal-to-Noise Ratio Constrained Channels
Feedback Stabilization over Signal-to-Noise Ratio Constrained Channels Julio H. Braslavsky, Rick H. Middleton and Jim S. Freudenberg Abstract There has recently been significant interest in feedback stabilization
More informationPerformance of an Adaptive Algorithm for Sinusoidal Disturbance Rejection in High Noise
Performance of an Adaptive Algorithm for Sinusoidal Disturbance Rejection in High Noise MarcBodson Department of Electrical Engineering University of Utah Salt Lake City, UT 842, U.S.A. (8) 58 859 bodson@ee.utah.edu
More informationModule 09 From s-domain to time-domain From ODEs, TFs to State-Space Modern Control
Module 09 From s-domain to time-domain From ODEs, TFs to State-Space Modern Control Ahmad F. Taha EE 3413: Analysis and Desgin of Control Systems Email: ahmad.taha@utsa.edu Webpage: http://engineering.utsa.edu/
More informationReduced-order Model Based on H -Balancing for Infinite-Dimensional Systems
Applied Mathematical Sciences, Vol. 7, 2013, no. 9, 405-418 Reduced-order Model Based on H -Balancing for Infinite-Dimensional Systems Fatmawati Department of Mathematics Faculty of Science and Technology,
More informationAchievable performance of multivariable systems with unstable zeros and poles
Achievable performance of multivariable systems with unstable zeros and poles K. Havre Λ and S. Skogestad y Chemical Engineering, Norwegian University of Science and Technology, N-7491 Trondheim, Norway.
More informationCONTROL DESIGN FOR SET POINT TRACKING
Chapter 5 CONTROL DESIGN FOR SET POINT TRACKING In this chapter, we extend the pole placement, observer-based output feedback design to solve tracking problems. By tracking we mean that the output is commanded
More information6.245: MULTIVARIABLE CONTROL SYSTEMS by A. Megretski. Solutions to Problem Set 1 1. Massachusetts Institute of Technology
Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science 6.245: MULTIVARIABLE CONTROL SYSTEMS by A. Megretski Solutions to Problem Set 1 1 Problem 1.1T Consider the
More information16.31 Fall 2005 Lecture Presentation Mon 31-Oct-05 ver 1.1
16.31 Fall 2005 Lecture Presentation Mon 31-Oct-05 ver 1.1 Charles P. Coleman October 31, 2005 1 / 40 : Controllability Tests Observability Tests LEARNING OUTCOMES: Perform controllability tests Perform
More informationOn Discrete-Time H Problem with a Strictly Proper Controller
On Discrete-Time H Problem with a Strictly Proper Controller Leonid Mirkin Faculty of Mechanical Engineering Technion IIT Haifa 3000, Israel E-mail: mersglm@txtechnionacil May 8, 1996 Abstract In this
More informationMultivariable Control. Lecture 03. Description of Linear Time Invariant Systems. John T. Wen. September 7, 2006
Multivariable Control Lecture 3 Description of Linear Time Invariant Systems John T. Wen September 7, 26 Outline Mathematical description of LTI Systems Ref: 3.1-3.4 of text September 7, 26Copyrighted
More informationẋ n = f n (x 1,...,x n,u 1,...,u m ) (5) y 1 = g 1 (x 1,...,x n,u 1,...,u m ) (6) y p = g p (x 1,...,x n,u 1,...,u m ) (7)
EEE582 Topical Outline A.A. Rodriguez Fall 2007 GWC 352, 965-3712 The following represents a detailed topical outline of the course. It attempts to highlight most of the key concepts to be covered and
More informationPerformance Limitations for Linear Feedback Systems in the Presence of Plant Uncertainty
1312 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 48, NO. 8, AUGUST 2003 Performance Limitations for Linear Feedback Systems in the Presence of Plant Uncertainty Graham C. Goodwin, Mario E. Salgado, and
More informationNorm invariant discretization for sampled-data fault detection
Automatica 41 (25 1633 1637 www.elsevier.com/locate/automatica Technical communique Norm invariant discretization for sampled-data fault detection Iman Izadi, Tongwen Chen, Qing Zhao Department of Electrical
More informationWEAK STRUCTURE AT INFINITY AND ROW-BY-ROW DECOUPLING FOR LINEAR DELAY SYSTEMS
KYBERNETIKA VOLUME 40 (2004), NUMBER 2, PAGES 181 195 WEAK STRUCTURE AT INFINITY AND ROW-BY-ROW DECOUPLING FOR LINEAR DELAY SYSTEMS Rabah Rabah and Michel Malabre We consider the row-by-row decoupling
More informationAnalysis of robust performance for uncertain negative-imaginary systems using structured singular value
8th Mediterranean Conference on Control & Automation Congress Palace Hotel, Marrakech, Morocco June 3-5, 00 Analysis of robust performance for uncertain negative-imaginary systems using structured singular
More informationEmpirical Gramians and Balanced Truncation for Model Reduction of Nonlinear Systems
Empirical Gramians and Balanced Truncation for Model Reduction of Nonlinear Systems Antoni Ras Departament de Matemàtica Aplicada 4 Universitat Politècnica de Catalunya Lecture goals To review the basic
More informationHANKEL-NORM BASED INTERACTION MEASURE FOR INPUT-OUTPUT PAIRING
Copyright 2002 IFAC 15th Triennial World Congress, Barcelona, Spain HANKEL-NORM BASED INTERACTION MEASURE FOR INPUT-OUTPUT PAIRING Björn Wittenmark Department of Automatic Control Lund Institute of Technology
More informationCONTROL SYSTEMS, ROBOTICS AND AUTOMATION CONTENTS VOLUME VII
CONTENTS VOLUME VII Control of Linear Multivariable Systems 1 Katsuhisa Furuta,Tokyo Denki University, School of Science and Engineering, Ishizaka, Hatoyama, Saitama, Japan 1. Linear Multivariable Systems
More informationAn Observation on the Positive Real Lemma
Journal of Mathematical Analysis and Applications 255, 48 49 (21) doi:1.16/jmaa.2.7241, available online at http://www.idealibrary.com on An Observation on the Positive Real Lemma Luciano Pandolfi Dipartimento
More informationOutput Adaptive Model Reference Control of Linear Continuous State-Delay Plant
Output Adaptive Model Reference Control of Linear Continuous State-Delay Plant Boris M. Mirkin and Per-Olof Gutman Faculty of Agricultural Engineering Technion Israel Institute of Technology Haifa 3, Israel
More informationMultivariable MRAC with State Feedback for Output Tracking
29 American Control Conference Hyatt Regency Riverfront, St. Louis, MO, USA June 1-12, 29 WeA18.5 Multivariable MRAC with State Feedback for Output Tracking Jiaxing Guo, Yu Liu and Gang Tao Department
More informationDELAY-DIFFERENTIAL systems arise in the study
438 IEEE TRANSACTIONS ON AUTOMATIC CONTROL On Feedback Stabilizability of Linear Systems With State Input Delays in Banach Spaces Said Hadd Qing-Chang Zhong, Senior Member, IEEE Abstract The feedback stabilizability
More informationSpectral factorization and H 2 -model following
Spectral factorization and H 2 -model following Fabio Morbidi Department of Mechanical Engineering, Northwestern University, Evanston, IL, USA MTNS - July 5-9, 2010 F. Morbidi (Northwestern Univ.) MTNS
More informationModule 03 Linear Systems Theory: Necessary Background
Module 03 Linear Systems Theory: Necessary Background Ahmad F. Taha EE 5243: Introduction to Cyber-Physical Systems Email: ahmad.taha@utsa.edu Webpage: http://engineering.utsa.edu/ taha/index.html September
More informationA Novel Integral-Based Event Triggering Control for Linear Time-Invariant Systems
53rd IEEE Conference on Decision and Control December 15-17, 2014. Los Angeles, California, USA A Novel Integral-Based Event Triggering Control for Linear Time-Invariant Systems Seyed Hossein Mousavi 1,
More informationEECE Adaptive Control
EECE 574 - Adaptive Control Overview Guy Dumont Department of Electrical and Computer Engineering University of British Columbia Lectures: Thursday 09h00-12h00 Location: PPC 101 Guy Dumont (UBC) EECE 574
More informationIntermediate Process Control CHE576 Lecture Notes # 2
Intermediate Process Control CHE576 Lecture Notes # 2 B. Huang Department of Chemical & Materials Engineering University of Alberta, Edmonton, Alberta, Canada February 4, 2008 2 Chapter 2 Introduction
More informationDescriptor system techniques in solving H 2 -optimal fault detection problems
Descriptor system techniques in solving H 2 -optimal fault detection problems Andras Varga German Aerospace Center (DLR) DAE 10 Workshop Banff, Canada, October 25-29, 2010 Outline approximate fault detection
More informationChapter 30 Minimality and Stability of Interconnected Systems 30.1 Introduction: Relating I/O and State-Space Properties We have already seen in Chapt
Lectures on Dynamic Systems and Control Mohammed Dahleh Munther A. Dahleh George Verghese Department of Electrical Engineering and Computer Science Massachuasetts Institute of Technology 1 1 c Chapter
More information